GLM-4
Zhipu AI's bilingual large language model offering conversational chat, coding assistance, and long-context reasoning in both Chinese and English.
About
GLM-4 is a series of open-weight large language models developed by Zhipu AI, accessible through the ChatGLM chat interface at chatglm.cn and via API. The models are designed for smooth bilingual conversation in Chinese and English and perform well on coding benchmarks, front-end generation, and multi-turn reasoning tasks. Newer releases in the series, including GLM-4.6 and GLM-4.7, feature expanded context windows up to 200K tokens and improved agentic capabilities such as preserved reasoning chains across turns. Lighter Flash variants are available for free with a daily token quota, while heavier models and a coding subscription plan are offered at paid tiers.